On Sun, 13 May 2007, [EMAIL PROTECTED] wrote:
>   1 -- Why do I get Transmission curve greator than 1 if it is normalized
> to the incident field?

Probably you're not normalizing correctly.  To normalize, you generally 
need to do a *separate* calculation with just a straight waveguide (or 
whatever your structure is) and put a flux plane at the *end* of the 
waveguide to measure the incident power.

(If your transmission exceeds 1 by only a couple percent, it may be that 
you need to increase the PML thickness).

>   2 -- Why do I get a strange transmission curve for the 2-D photonic
> crystal shown below the curve. It seems to be a linear curve
> like the following:
>                            |     /
>                            |    /
>                            |   /
>                            |  /
>                            ------------
>
>  o o o o o o o o o o
>   o o o o o o o o o o
>  o o o o o o o o o o
>   o o o o o o o o o o
>          o o o
>   o o o o o o o o o o
>  o o o o o o o o o o o
>   o o o o o o o o o o
>  o o o o o o o o o o o

The transmission in this case should be very small for all the frequencies 
in the gap (since the field is exponentially decaying in the middle, if I 
read your diagram correctly, and there is no resonant cavity).  If you get 
transmission that is not small, probably you screwed up the normalization 
(see above).

>   3 -- How to normalize a waveguides transmission curve if it has no
> defect in it, because for normalization we do two runs i.e. one with
> the defect or some structure inside it and than with no defect or
> structure inside it like the following structure.
>
>  o o o o o o o o o o
>   o o o o o o o o o o
>  o o o o o o o o o o
>   o o o o o o o o o o
>                    < -- waveguide with no defect in it?
>   o o o o o o o o o o
>  o o o o o o o o o o o
>   o o o o o o o o o o
>  o o o o o o o o o o o

The question is, what are you trying to compute?  When you measure 
"transmission",  you are really measuring
        (power in X) / (power in Y)
and it is up to you to decide what X and Y are.

If you have just a straight waveguide, it's not clear what you're trying 
to compute the transmission with respect to.

Steven

PS. Note that Meep (really, Scheme) provides many ways to write loops to 
make lots of objects.  If you have a long list of objects in your .ctl 
file you should probably be using a loop of some sort instead.  For 
example, you can use geometric-objects-duplicates to duplicate a list of 
objects (i.e. so that you can nest duplication to create 2d lattices).

_______________________________________________
meep-discuss mailing list
[email protected]
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss

Reply via email to